Human hearing beats the Fourier uncertainty principle

[ciamara]

Hi all,

This article is making its way around engineering circles (and others):
http://m.phys.org/news/2013-02-human-fourier-uncertainty-principle.html

Finally some empirical evidence that we are not all crazy and we can actually hear beyond 20kHz. I have always said that while it is difficult to hear audio outside the 20Hz to 20kHz range, we can perceive it. Production and engineering are by far the most important when it comes to perceived sound quality, but it is clear that redbook (16bit/44kHz) just doesn't cut it. Long live high resolution digital audio!

 

[DonH50]

There are several other threads discussing this study already... See e.g. http://www.whatsbestforum.com/showthread.php?9582-Math-Unraveled And the study is in some dispute in scientific circles as to the methodology and what is really proves. As to the importance of reproducing frequencies above our hearing limits, I always thought that had more to do with mixing products that fall back in-band, but I am not an expert in anatomy/biology/physiology/whatever.

 

[ciamara]

Sorry! Didn't realize it had been discussed elsewhere. As for frequencies, many DACs and other devices in the studio involve filters that are designed to move noise into a spectrum that was thought to be inaudible. Now, it seems -- if we want to believe the results of this article -- these filters need to be revisited.

 

[mep]

Hi all,

This article is making its way around engineering circles (and others):
http://m.phys.org/news/2013-02-human-fourier-uncertainty-principle.html

Finally some empirical evidence that we are not all crazy and we can actually hear beyond 20kHz. I have always said that while it is difficult to hear audio outside the 20Hz to 20kHz range, we can perceive it. Production and engineering are by far the most important when it comes to perceived sound quality, but it is clear that redbook (16bit/44kHz) just doesn't cut it. Long live high resolution digital audio!

Like all statements that are considered to be inflammatory on this forum, you should prepare by buying some asbestos underwear in order to keep your posterior safe from bursting into flames. There are some people on this forum who would swear that MP3s can't be distinguished from hearing the BSO live at Symphony Hall (I'm slightly exaggerating, but not by much). There are some others who would swear that you couldn't hear the difference between a low bit-rate MP3 and a microphone feed of the actual recording event. So when you start taking on RB CD and saying it "just doesn't cut it," you are approaching the digital holy grail for some people and you need to lace up your gloves.

 

[ciamara]

Ha ha ha! Fair point. I think I oversimplified, so I will try to stave off the raging inferno with a garden hose ... I think we can all agree that the engineering that goes into the recording matters most (after the musicians, of course). How material is recorded, what equipment is used to track and mix, how well the engineer(s) use(s) dynamic range compression, EQ, filtering, side-chain processing, reverb, effects processing etc. -- all these have a huge impact. Great engineers can make CDs sound amazing -- no question about it. And in the digital world, most engineers who care about sound quality track and mix at higher sample rates than 44kHz. (Pretty much everyone uses greater bit depths to provide more dynamic range.) The thinking is that operating at a higher sample rate than the intended output format ensures nothing is lost in the mathematics. For CDs the final mix is reduced down to 16/44 (usually with dithering/noise shaping etc.). I guess what I'm driving at is this ... If instead of reducing the mix down to 16/44, we kept more of the resolution, less filtering and noise shaping would have to happen. We could record at 24/384 and use 24/192 as the output format. And this already happens today.

I spent some time in a respected mastering facility last weekend, and they seem very focused on producing 24/176 and 24/192 mixes. Perhaps it is driven by the demand side of the equation, but the engineers I spoke with seemed to feel the higher sample rates provide more detail.

I was also able to hear the "iPod plugin" that converts a mix from full res to Apple's 256 kbps compressed format -- in real time. It can be engaged or disengaged on the fly. It was not easy to tell the difference right away, but after careful listening it was obvious. My initial reaction was that the compression codec is quite good. But on very good equipment in an acoustically neutral room (as you'd expect in a mastering facility), the difference was clearly apparent. Less detail, cymbals and high frequency detail seemed a bit off.

 

[Ethan Winer]

Finally some empirical evidence that we are not all crazy and we can actually hear beyond 20kHz.

No flames here, but that article proves nothing of the sort. It's about the frequency discrimination limits of software, not the highest frequency human ears can hear. So you all are in fact crazy! [Kidding]

--Ethan

 

[edorr]

I was also able to hear the "iPod plugin" that converts a mix from full res to Apple's 256 kbps compressed format -- in real time. It can be engaged or disengaged on the fly. It was not easy to tell the difference right away, but after careful listening it was obvious. My initial reaction was that the compression codec is quite good. But on very good equipment in an acoustically neutral room (as you'd expect in a mastering facility), the difference was clearly apparent. Less detail, cymbals and high frequency detail seemed a bit off.

Still, if a trained ear needs carefull listening on very good equipment, it is obvious that compressed audio formats are totally adequate for the vast majority on non critical listening by mainstream consumers. As opposed to say Blu Ray versus DVD, which is a night and day difference immediately visible to everyone using a $200 BR player and a $700 display.

 

[ciamara]

No flames here, but that article proves nothing of the sort. It's about the frequency discrimination limits of software, not the highest frequency human ears can hear. So you all are in fact crazy! [Kidding]

--Ethan

Thanks Ethan. I probably am a bit crazy. No question about it!